Effect of air-frying on formation and correlations of polycyclic aromatic hydrocarbons, acrylamide, and heterocyclic aromatic amines in foods and risk assessment

Abstract

The levels of acrylamide (AA), four polycyclic aromatic hydrocarbons (PAH4), and heterocyclic aromatic amines (HAAs) in 184 air-fried agricultural, fishery, and animal products were measured using GC–MS and UPLC-MS/MS. Among the tested samples, sea algae exhibited the highest levels of PAH4 and eight specific HAAs (HAA8), while root and tuber crops had the greatest amount of AA. Agricultural and fisheries products had higher levels of all three contaminants, while livestock products had an inverse correlation between PAH4 and HAA8. The margin of exposure in the Korean population is considered “unlikely a concern” for all samples for PAH4 and HAA8, however, that for AA in cereal, vegetable, and root and tuber crops is deemed “may be a concern”, with a value < 10,000 in all age groups. These findings suggest a need to evaluate dietary AA exposure in certain food categories and further research to minimize AA formation during air frying.

Graphic abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s10068-024-01639-4.

Introduction

Traditional cooking methods like grilling, roasting, smoking, and frying offer various advantages in food preparation, but direct contact with flames can sometimes lead to the formation of heat-induced contaminants (Trafialek and Kolanowski, 2014). Polycyclic aromatic hydrocarbons (PAHs), organic molecules composed of carbon and hydrogen atoms arranged in at least two condensed or fused aromatic ring structures, are produced by the pyrolysis of organic substances and incomplete combustion (Lee et al., 2019). Smoking and barbecuing also serve as sources of PAH contamination in food (Bansal and Kim, 2015). Given that PAHs are primarily lipophilic, they efficiently bioaccumulate in organic tissues via the food chain (Pérez-Cadahía et al., 2004). The International Agency for Research on Cancer (IARC) classified benzo[a]pyrene (B[a]P) as carcinogenic to humans (Group 1) and benzo[a]anthracene (B[a]A), benzo[b]fluoranthene (B[b]F), and chrysene (Chr) as possibly carcinogenic to humans (Group 2B) (IARC, 2023). The European Food Safety Authority (EFSA) recommends using the sum of the concentrations of four specific polycyclic aromatic hydrocarbons (PAH4)—B[a]P, Chr, B[a]A, and B[b]F—as a marker for overall PAH contamination (PAH4). This approach is considered more accurate than relying solely on B[a]P levels (EC, 2015).

Acrylamide (AA) is a highly water-soluble organic compound and one of the thermal decomposition products of triacylglycerol (Rannou et al., 2016). Additionally, AA is predominantly produced in carbohydrate-rich foods such as coffee, potato chips, bread, and cereal using high-temperature (over 120 °C) cooking processes and low moisture content (EFSA, 2015; Zhang et al., 2009). AA was classified as a probable human carcinogen by the IARC and can be damaging to the neurological system in both humans and animals through chronic exposure (Huang et al., 2011; IARC, 2023).

Heterocyclic aromatic amines (HAAs), which include at least one aromatic ring and one amino group in their structure, are hazardous compounds found in high concentrations in amino acids and creatine-containing foods (Adeyeye, 2020). Over 25 HAAs produced during cooking have the potential to cause mutations by converting into electrophiles and binding to DNA. Additionally, these chemicals can cause tumors in the reproductive organs of mice and rats (Alaejos and Afonso, 2011), and then, there are classified as possible carcinogens (Group 2) in IARC (IARC, 2023).

Air-fryers were recently introduced to limit the creation of compounds generated by heat and are commonly used in home cooking due to their ease in cooking a small volume of food. They are replacing deep-frying by adding a small amount of oil (Lee et al., 2020). The air fryer’s function is to keep the chamber temperature consistent while the applied oil on the food is dispersed by hot air convection as oil droplets, making contact with the food to transmit heat (Zaghi et al., 2019). The rise in health awareness is a factor contributing to the current popularity of air fryers, and consumers are paying more attention to what they eat and trying to eat healthier (Pérez-Pedraza et al., 2022).

PAHs, AA, and HAAs, considered carcinogenic and genotoxic compounds, share the characteristic of being produced through the Maillard reaction, where amino acids and reducing sugars react at high temperatures during prolonged cooking times (Adeyeye, 2020; Skog et al., 1998; Zhang et al., 2009). There has not been much research investigating the formation of these toxic compounds when different types of food are cooked in different ways utilizing an air fryer. Therefore, a study monitoring and evaluating the risk of these hazardous compounds produced by utilizing an air fryer is required. The purpose of this study is to determine the levels of PAH4, AA, and HAAs in foods prepared using the most popular air fryer recipes in South Korea and to assess the potential harm to the people in the country.

Materials and methods

Sample preparation

Each of the samples was prepared using a Phillips air fryer (HD9270/90, Eindhoven, Netherlands). All food ingredients were sourced from regional markets. Cooking methods from recipe books, social media, and influential blogs in South Korea were selected for preparing foods with the air fryer. In total, 233 food samples were analyzed, and for each sample, a duplicate analysis of PAHs, AA, and HAAs was conducted. The categorization of food groups was based on the 6th Korean National Health and Nutrition Examination Survey (KNHANES-VI) conducted by the Korea Disease Control and Prevention Agency (KDCA) from 2013 to 2015 (MOHW, 2013). It allowed for the classification of samples into 10 categories. This categorization was done to facilitate future risk assessment studies that may utilize the KNHANES-VI data. The specific food items within each category were selected to represent commonly consumed foods in the Korean diet, ensuring that the study results would be relevant to the general population. Factors such as the potential for these food items to contain the harmful substances of interest were also considered, based on previous studies and the understanding of their preparation methods. The recipes with details on cooking time, temperature, and seasoning were provided in Supplementary Table 1. Sample categories were classified into cereal (n = 22), vegetable (n = 24), root & tuber crops (n = 10), mushroom (n = 8), nut & seed (n = 4), fruit (n = 1), fish & shellfish (n = 41), sea algae (n = 7), meat (n = 66), and egg (n = 1). After cooking, the samples were homogenized and stored at − 20 °C. Analyses were conducted twice on each sample.

Standards of PAHs, AA, and HAAs

To analyze the contents of PAH4 comprising B[a]P, B[a]A, B[b]F, and Chr, the standard solution (EPA 525 PAH Mix A) and each internal standards (B[a]A-d₁₂, B[a]P-d₁₂, B[b]F-d₁₂, and Chr-d₁₂) were purchased from Sigma Aldrich (Saint Louis, MO, USA). The standard solution and internal standard solution (¹³C₃-acrylamide) for AA were purchased from Sigma-Aldrich (St Louis, MO, USA) and Cambridge Isotope Laboratories, Inc. (Andover, MA, USA), respectively. To analyze the contents of HAAs, authentic standards of 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), 2-amino-3,4-dimethyl-3H-imidazo[4,5-f]quinoline (MeIQ), 2-amino-1-methyl-6-phenylimidazo[4,5-b] pyridine (PhIP), 2-amino-9H-pyrido[2,3-b]indole (AαC), harman, and norharman were purchased from Cayman Chemical (Ann Arbor, MI, USA) or Sigma-Aldrich (Saint Louis, MO, USA). The rest of HAA standards (2-amino-3,8-dimethylimidazo[4,5-f]quinoline quinoxaline (MeIQx), 2-amino-3-methyl-9H-pyrido[2,3-b] indole (MeAαC), 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole acetate (Trp-P-1), and 3-amino-1-methyl-5H-pyrido[4,3-b]indole acetate (Trp-P-2)) and internal standards for HAAs (IQ-d₃, MeIQ-d₃, MeIQx-d₃, PhIP-d₃, MeAαC-d₃, Trp-P-2-¹³C₂, harman-d₃, and norharman-d₇) were purchased from Tronto Research Chemical (Toronto, Ontario, Canada). HAA8 was named after the eight HAAs with known carcinogenicity (IQ, MeIQ, MeIQx, PhIP, AαC, MeAαC, Trp-P-1, and Trp-P-2), and HAA10 was named by the inclusion of two non-carcinogens (Harman and norharman) in this study.

Method validation

For method validation, the correlation coefficient (r²), method limits of detection (LOD), and method limits of quantification (LOQ) in Table 1 were obtained using as follows: For PAH4, using the salmons with alkali saponification as a sample matrix considering matrix effects, 6 points (0.5–20 µg/L) of the standard solution with 100 µg/L of the internal standard (diluted in dichloromethane (DCM; Honeywell, Charlotte, NC, USA)) were utilized to obtain the calibration curve. The calibration curve for AA was obtained using salmon as a sample matrix and 8 points (0–20 µg/L) of the standard solution with 200 µg/L of the internal standard (diluted in 0.1% formic acid (Merk Chemical, Darmstadt, Germany)). For HAAs, it was utilized the pork loins, as a sample matrix, and 7 points (0.5–50 µg/L) of standard solution with 1 µg/L of the internal standard (diluted in methanol (Honeywell, Charlotte, NC, USA)). In addition, Accuracy and precision are shown in Supplementary Tables 2 to 4.

Table 1.

Correlation coefficient (r²), limits of detection (LOD), and limits of quantification (LOQ) of four specific polycyclic aromatic hydrocarbons (PAH4), acrylamide (AA), and eight specific heterocyclic aromatic amines (HAA8)

Groups	Compounds	Abbreviation	Quantitative ion (m/z)	Qualitative ion (m/z)	Linearity (r2)	LOD) (µg/kg)	LOQ (µg/kg)
PAH4	Benz[a]anthracene	B[a]A	228	226, 229	0.999	0.09	0.27
	Chrysene	Chry	228	226, 229	0.999	0.07	0.21
	Benzo[b]fluoranthene	B[b]F	252	250, 253	0.999	0.09	0.28
	Benzo[a]pyrene	B[a]P	252	250, 253	0.999	0.08	0.24
AA	Acrylamide	AA	72 → 55	72 → 27	0.999	0.63	2.08
HAA8	2-Amino-3-methylimidazo[4,5-f]quinoline	IQ	184	157, 131	0.997	0.03	0.10
	2-Amino-3,4-dimethyl-3H-imidazo[4,5-f]quinoline	MeIQ	198	145, 172	0.999	0.05	0.16
	2-Amino-3,8-dimethylimidazo[4,5-f]quinoxaline	MeIQx	199	173, 146	0.999	0.04	0.13
	2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine	PhIP	210	183, 157	0.999	0.05	0.16
	2-Amino-9H-pyrido[2,3-b]indole	AαC	167	157, 140	0.999	0.04	0.13
	2-Amino-3-methyl-9H-pyrido[2,3-b]indole	MeAαC	181	129, 154	0.999	0.04	0.11
	3-Amino-1,4-dimethyl-5H-pyrido[4,3-b]indole acetate	Trp-P-1	195	168, 141	0.997	0.05	0.14
	3-Amino-1-methyl-5H-pyrido[4,3-b]indole acetate	Trp-P-2	181	154, 130	0.997	0.03	0.09
	1-Methyl-9H-pyrido[3,4-b]indole	Harman	115	168, 142	0.998	0.05	0.14
	9H-pyrido[3,4-b]indole	Norharman	115	89, 142	0.999	0.04	0.11

Extraction and purification of samples for PAHs

All samples were used to extract PAHs after alkali saponification as follows: In a round-bottomed flask containing approximately 10 g of homogenized sample, 100 mL of 1 M KOH solution (Showa Denko, Minato, Tokyo, Japan) in ethanol (Burdick & Jackson, Muskegon, MI, USA) and 1 mL of 100 μg/L internal standard solution of B[a]A-d₁₂, B[a]P-d₁₂, B[b]F-d₁₂, and Chr-d₁₂ in DCM were added. After alkali saponification by heating for 3 h at 80 °C in a water bath (Daihan Scientific, Wonju, Gangwon, Republic of Korea), the flask was rapidly cooled in ice water and 50 mL of n-hexane was added.

The extracted sample was then transferred to a separatory funnel using filter paper. The sample was mixed with 50 mL of n-hexane (Honeywell, Charlotte, NC, USA): ethanol (Burdick & Jackson, Muskegon, MI, USA) (1:1, v/v) and 50 mL of distilled water. After shaking the separatory funnel for 5 min at 300 rpm using a shaker (Changshin Science, Jongno, Seoul, Republic of Korea), the solvents were separated into the organic phase and distilled water layer. The organic phases were collected, before repeating the extraction step twice by adding 50 mL of n-hexane to the distilled water layer. The collected organic phases were added with 100 mL of distilled water to separate the water layer, and only the organic phases were used for the next steps. The organic phases were dehydrated with 15 g of anhydrous sodium sulfate in round-bottomed flasks to eliminate water completely. The dehydrated solvents were concentrated by a rotary evaporator (EYELA, Koishikawa Bunkyo, Tokyo, Japan) in a water bath at 40 °C until the final volumes were 2 mL. To dissolve the concentrated solvents, 5 mL of n-hexane and 15 mL of n-hexane: DCM (3:1, v/v) were added to the flasks.

For purification, the solvents, obtained using solid-phase extraction (SPE) cartridges (Bond Elut SI; Agilent Technologies, Santa Clara, CA, USA) preconditioned with 5 mL of DCM and 15 mL of n-hexane, were concentrated with nitrogen (N₂) gas at 40 °C. Finally, after dissolving these residues by adding 1 mL of DCM, the dissolved solvents were purified with PTFE membrane syringe filters (Whatman, Maidstone, Kent, UK) and analyzed by GC–MS (7890B/5977B; Agilent Technologies, Santa Clara, CA, USA) with a Zebron ZB-PAH-SeleCT column (40 × 0.18 mm × 0.14 μm; Phenomenex, Torrance, CA, USA). The oven temperature was initially set at 45 °C for 0.8 min, increased to 200 °C at 45 °C/min, then to 265 °C at 3 °C/min with a 5-min hold, further increased to 270 °C at 1 °C/min, and finally raised to 320 °C at 10 °C/min with a 15-min hold. Injection mode in splitless mode (1 µL) at 320 °C was employed, and the mass spectrometer operated in Electron Impact at 70 eV using Selected Ion Monitoring mode. Helium served as the carrier gas with a constant flow rate of 1 mL/min.

Extraction and purification of samples for AA

AA in samples was extracted with water following the addition of 1 mL of internal standard. Approximately 1 g of homogenized samples was weighed in 50 mL conical tubes. To the sample tubes, 1 mL of 200 µg/L of an internal standard solution and 9 mL of water were added. Each sample tube was shaken for 20 min at 250 rpm and centrifuged for 5 min at 3500 rpm. The supernatants were purified with a PVDF syringe of 0.45 µm (Futecs, Daejeon, Republic of Korea). The sample purification was conducted with two SPE cartridges (Strata-X (Phenomenex, Torrance, CA, USA) and Bond Elut AccuCAT (Agilent Technologies, Santa Clara, CA, USA)) in series. The Strata-X SPE cartridge was first rinsed with 3.5 mL of methanol and 3.5 mL of water. After that, 1.5 mL of purified samples were eluted into this cartridge followed by 0.5 mL of water, and the eluent was discarded. Then, an additional 1.5 mL of water was loaded onto the Strata-X SPE cartridge, and the eluent was collected. Bond Elut AccuCAT SPE cartridges were conditioned with 2.5 mL of methanol and 2.5 mL of water. Finally, 0.5 mL of the eluent collected from the Strata-X cartridge was passed through the Bond Elut AccuCAT SPE cartridge and discarded. The Bond Elut AccuCAT SPE cartridge was then loaded with the remaining 1 mL of eluent from the Strata-X cartridge, and filtered samples were analyzed by HPLC–MS/MS with a UHPLC Nexera X2 system (Shimadzu, Kyoto, Japan) coupled to a triple quadrupole mass spectrometer (TQMS8040 MS/MS; Shimadzu, Kyoto, Japan). A Kinetex Polar C18 100 Å column (2.1 × 150 mm, particle size 2.6 µm) (Phenomenex, Torrance, California, USA) maintained at 26 °C facilitated the separation of AA. The isocratic mobile phase (0.2% acetic acid and 5% methanol in distilled water) was delivered at a flow rate of 0.3 mL/min, and the injection volume was 10 µL. Electrospray ionization with a positive mode was employed.

Extraction and purification of samples for HAAs

All samples underwent SPE to extract HAAs. Approximately 3 g of homogenized samples were weighed into 50 mL conical tubes. The samples were then mixed with 15 mL of 1 M NaOH solution (Sigma Aldrich, Saint Louis, MO, USA) and 20 µL of a 1 µg/mL internal standard solution. After the addition of 15 mL of acetonitrile (J.T.Baker, Radnor, Pennsylvania, USA) to the mixtures, the solutions were shaken for 10 min at 700 rpm using a shaker (Hankuk S&I, Hwaseong, Gyeonggi-do, Republic of Korea). Subsequently, the mixtures were centrifuged at 4 °C for 25 min at 4,000 rpm. The supernatants were collected, and a second round of extractions was performed by adding 10 mL of acetonitrile to the pellets. The extracts were loaded onto a Chem Elut cartridge (Agilent Technologies, Santa Clara, CA, USA) and eluted with DCM: ethyl acetate (Sigma Aldrich, Saint Louis, MO, USA) (8:2, v/v). The eluents were evaporated and the residues were dissolved in 2 mL of methanol.

Oasis HLB SPE cartridges (6 cc/200 mg; Waters, Milford, MA, USA)) were preconditioned with 4 mL of water and 4 mL of methanol. Subsequently, 2 mL of the extracts were loaded onto the SPE cartridge, and HAAs were eluted with 8 mL of methanol. A nitrogen concentrator (MG-2200, EYELA, Tokyo, Japan) was emplyed to evaporate the eluents under N₂ gas until the final volumes reached 0.5 mL. Finally, the extracts underwent purification with 0.22 μm nylon filters (Whatman, Maidstone, Kent, UK), and the filtered samples were analyzed by UPLC (1290; Agilent Technologies, Santa Clara, CA, USA) coupled with a 6470B triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA). HAAs were separated on an ACQUITY UPLC BEH C18 column (2.1 × 100 mm, 1.7 μm; Waters, Milford, MA, USA) maintained at 40 °C. A gradient of 30 mM ammonium formate (A) and methanol (B) was used, following this profile: 10–15% (B), 0–2.5 min; 15–20% (B), 2.5–4.0 min; 20% (B), 4.0–7.5 min; 20–25% (B), 7.5–9.5 min; 25% (B), 9.5–11.0 min; 25–35% (B), 11.0–13.0 min; 35–95% (B), 13.0–15.0 min; 95% (B), 15.0–23.0 min; 95–10% (B), 23.0–30.0 min. The flow rate was 0.3 mL/min, and the injection volume was 5 µL. The source temperature and nebulizer gas pressure were maintained at 250 °C and 30 psi, respectively.

Dietary exposure and margin of exposure

Because many PAHs cause cancer in different ways, the TEQ (Toxic Equivalency) method may not be ideal for assessing their overall risk (EFSA, 2008). Individual toxicological data for each PAH provides a more accurate approach. Therefore, to calculate the total concentration of four specific PAHs (TC_PAH4), Eq. (1) was used to combine the measured concentrations of each PAH. The dietary exposure and margin of exposure were assessed in this study using the lower bound of the benchmark dose confidence interval (BMDL) for PAH4, AA, PhIP, and HAA8 (EFSA, 2022). Consequently, the mean and 95th percentile (P95) dietary exposure of PAH4, PhIP, HAA8, and AA, along with mean and P95 margin of exposure (MOE) of PAH4, AA, PhIP, and HAA8 were estimated. In contrast, HAA8, PhIP, and AA were used as the means concentrations in Table 2.

$$T{C}_{PAH4i}=\sum _{i=1}^n{C}_i\left(\frac{\mathit{ng}}{g}\right)$$ 1

where, TC_PAH4 is the total concentration of the ith individual congener of PAH4, and Ci is the measured concentration for the ith individual congener of PAH4. Values below the LOD in Tables 3 and 4 were substituted following WHO (1995) as follows: For PAH4 values below the LOD, replacement with lower bound (LB) and upper bound (UB) values occurred when the fraction of values below the LOD was between 60 and 80%. The LB and UB values utilized were zero and LOD, respectively. Values below the LOD were replaced using half of the LOD when the fraction of values below the LOD was less than 60%. HAA8, PhIP, and AA values below the LOD were replaced with half of the LOD.

Table 2.

Concentrations of PAH4, AA, and HAAs detected in foods cooked using the air fryer

Categories		Agricultural products (n = 69)						Fishery products (n = 48)		Livestock products (n = 67)
		Cereals (n = 22)	Vegetables (n = 24)	Root & tuber crops (n = 10)	Mushrooms (n = 8)	Nuts & seeds (n = 4)	Fruits (n = 1)	Fish & shellfish (n = 41)	Sea algae (n = 7)	Meat (n = 66)	Eggs (n = 1)
B[a]A	Detected sample (%)d	1 (5%)	2 (8%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	10 (24%)	1 (14%)	7 (11%)	0 (0%)
	Mean (µg/kg)e	0.02	0.02	0.00	0.00	0.00	0.00	0.09	0.03	0.03	0.00
	Min–max (µg/kg)	0.00–0.35	0.00–0.22	0.00	0.00	0.00	0.00	0.00–1.36	0.00–0.19	0.00–0.86	0.00
Chr	Detected sample (%)	1 (5%)	5 (21%)	1 (10%)	0 (0%)	1 (25%)	0 (0%)	19 (46%)	4 (57%)	5 (8%)	0 (0%)
	Mean (µg/kg)	0.02	0.04	0.01	0.00	0.03	0.00	0.22	0.13	0.03	0.00
	Min–max (µg/kg)	0.00–0.43	0.00–0.26	0.00–0.11	0.00	0.00–0.12	0.00	0.00–2.49	0.00–0.44	0.00–0.97	0.00
B[b]F	Detected sample (%)	1 (5%)	1 (4%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	7 (17%)	1 (14%)	1 (2%)	0 (0%)
	Mean (µg/kg)	0.01	0.00	0.00	0.00	0.00	0.00	0.05	0.02	0.00	0.00
	Min–max (µg/kg)	0.00–0.21	0.00–0.11	0.00	0.00	0.00	0.00	0.00–0.72	0.00–0.12	0.00–0.15	0.00
B[a]P	Detected sample (%)	2 (9%)	2 (8%)	1 (10%)	0 (0%)	0 (0%)	0 (0%)	3 (7%)	4 (57%)	5 (8%)	0 (0%)
	Mean (µg/kg)	0.07	0.01	0.01	0.00	0.00	0.00	0.02	0.38	0.02	0.00
	Min–max (µg/kg)	0.00–1.03	0.00–0.10	0.00–0.12	0.00	0.00	0.00	0.00–0.62	0.00–1.55	0.00–0.58	0.00
PAH4a	Detected sample (%)	2 (9%)	6 (25%)	2 (20%)	0 (0%)	1 (25%)	0 (0%)	23 (56%)	6 (86%)	11 (17%)	0 (0%)
	Mean (µg/kg)	0.12	0.07	0.02	0.00	0.03	0.00	0.38	0.56	0.08	0.00
	Min–max (µg/kg)	0.00–1.53	0.00–0.59	0.00–0.12	0.00	0.00–0.12	0.00	0.00–5.20	0.00–1.55	0.00–2.16	0.00
AA	Detected sample (%)	22 (100%)	24 (100%)	10 (100%)	8 (100%)	4 (100%)	1 (100%)	39 (95%)	7 (100%)	66 (100%)	0 (0%)
	Mean (µg/kg)	16.52	501.74	583.89	15.43	359.52	3.07	4.77	92.31	7.32	0.00
	Min–max (µg/kg)	1.72–79.87	1.14–5312.65	1.07–2645.76	2.30–83.25	4.49–1362.85	3.07–3.07	0.00–36.42	6.71–373.39	0.66–38.65	0.00
IQ	Detected sample (%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)
	Mean (µg/kg)	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
	Min–max (µg/kg)	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
MeIQ	Detected sample (%)	0 (0%)	0 (0%)	1 (10%)	2 (25%)	0 (0%)	0 (0%)	2 (5%)	0 (0%)	0 (0%)	0 (0%)
	Mean (µg/kg)	0.00	0.00	0.06	0.14	0.00	0.00	0.01	0.00	0.00	0.00
	Min–max (µg/kg)	0.00	0.00	0.00–0.56	0.00–0.65	0.00	0.00	0.00–0.50	0.00	0.00	0.00
MeIQx	Detected sample (%)	1 (5%)	1 (4%)	0 (0%)	0 (0%)	0 (0%)	1 (100%)	4 (10%)	1 (14%)	16 (24%)	0 (0%)
	Mean (µg/kg)	0.00	0.01	0.00	0.00	0.00	0.09	0.01	0.04	0.04	0.00
	Min–max (µg/kg)	0.00–0.06	0.00–0.12	0.00	0.00	0.00	0.09–0.09	0.00–0.30	0.00–0.31	0.00–0.61	0.00
PhIP	Detected sample (%)	10 (45%)	11 (46%)	4 (40%)	8 (100%)	2 (50%)	1 (100%)	32 (78%)	7 (100%)	62 (94%)	1 (100%)
	Mean (µg/kg)	0.36	0.32	0.36	1.29	0.34	0.32	0.65	1.66	0.49	0.76
	Min–max (µg/kg)	0.00–1.71	0.00–2.32	0.00–2.75	0.32–3.42	0.00–1.18	0.32–0.32	0.00–8.25	0.10–5.85	0.00–1.62	0.76–0.76
AαC	Detected sample (%)	2 (9%)	6 (25%)	0 (0%)	4 (50%)	0 (0%)	0 (0%)	8 (20%)	6 (86%)	14 (21%)	0 (0%)
	Mean (µg/kg)	0.02	0.06	0.00	0.40	0.00	0.00	0.14	0.35	0.10	0.00
	Min–max (µg/kg)	0.00–0.28	0.00–0.48	0.00	0.00–1.84	0.00	0.00	0.00–3.90	0.00–0.49	0.00–1.39	0.00
MeAαC	Detected sample (%)	6 (27%)	3 (13%)	0 (0%)	4 (50%)	0 (0%)	0 (0%)	9 (22%)	7 (100%)	36 (55%)	1 (100%)
	Mean (µg/kg)	0.02	0.02	0.00	0.14	0.00	0.00	0.09	1.08	0.06	0.15
	Min–max (µg/kg)	0.00–0.17	0.00–0.31	0.00	0.00–0.46	0.00	0.00	0.00–2.49	0.16–2.62	0.00–0.25	0.15–0.15
Trp-P-1	Detected sample (%)	4 (18%)	2 (8%)	0 (0%)	2 (25%)	0 (0%)	0 (0%)	10 (24%)	3 (43%)	26 (39%)	1 (100%)
	Mean (µg/kg)	0.03	0.01	0.00	0.03	0.00	0.00	0.05	0.43	0.06	0.18
	Min–max (µg/kg)	0.00–0.22	0.00–0.14	0.00	0.00–0.15	0.00	0.00	0.00–0.60	0.00–1.11	0.00–0.50	0.18–0.18
Trp-P-2	Detected sample (%)	1 (5%)	1 (4%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	1 (2%)	3 (43%)	3 (5%)	0 (0%)
	Mean (µg/kg)	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.38	0.00	0.00
	Min–max (µg/kg)	0.00–0.05	0.00–0.03	0.00	0.00	0.00	0.00	0.00–0.11	0.00–1.24	0.00–0.08	0.00
Harman	Detected sample (%)	22 (100%)	23 (96%)	9 (90%)	8 (100%)	4 (100%)	1 (100%)	41 (100%)	7 (100%)	66 (100%)	1 (100%)
	Mean (µg/kg)	11.51	3.50	1.95	1.84	4.19	1.27	6.18	8.09	8.97	0.87
	Min–max (µg/kg)	0.30–118.09	0.00–30.14	0.00–14.52	0.26–5.29	0.64–10.69	1.27–1.27	0.28–69.65	1.77–20.12	0.26–139.45	0.87–0.87
Norharman	Detected sample (%)	22 (100%)	24 (100%)	10 (100%)	8 (100%)	4 (100%)	1 (100%)	41 (100%)	7 (100%)	66 (100%)	1 (100%)
	Mean (µg/kg)	34.57	21.26	9.67	22.75	10.97	1.30	34.90	20.69	36.63	2.39
	Min–max (µg/kg)	0.30–81.05	0.28–95.51	0.22–41.46	1.12–128.28	1.42–22.00	1.30–1.30	0.64–206.55	4.89–87.62	1.03–100.55	2.39–2.39
HAA8b	Detected sample (%)	11 (50%)	13 (54%)	4 (40%)	8 (100%)	2 (50%)	1 (100%)	33 (80%)	7 (100%)	62 (94%)	1 (100%)
	Mean (µg/kg)	0.43	0.42	0.42	2.00	0.34	0.41	0.95	3.91	0.75	1.10
	Min–max (µg/kg)	0.00–2.22	0.00–2.80	0.00–3.31	0.32–6.35	0.00–1.18	0.41–0.41	0.00–15.54	0.26–9.47	0.00–2.60	1.10–1.10
HAA10c	Detected sample (%)	22 (100%)	24 (100%)	10 (100%)	8 (100%)	4 (100%)	1 (100%)	41 (100%)	7 (100%)	66 (100%)	1 (100%)
	Mean (µg/kg)	46.50	25.18	12.05	26.59	15.50	2.98	42.04	32.71	46.35	4.35
	Min–max (µg/kg)	0.62–153.69	0.80–102.20	0.22–41.64	2.08–128.89	2.97–33.86	2.98–2.98	0.98–278.97	7.00–109.64	2.22–185.07	4.35–4.35

^aThe sum of four PAHs (B[a]A, Chr, B[b]F, and B[a]P)

^bThe sum of eight HAAs (IQ, MeIQ, MeIQx, PhIP, AαC, MeAαC, Trp-P-1, and, Trp-P-2)

^cThe sum of ten HAAs (HAA8, Harman, and Norharman)

^dNumber of detected samples (proportion of samples detected)

^eValue of below LOD were substituted zero

Table 3.

Estimated dietary exposure of PAH4, AA, PhIP, and HAA8

Categories	Age group (years)	PAH4						AA			PhIP			HAA8
		TCPAH4a (µg/kg)		Mean dietary exposureb		p95 Dietary exposure		Mean AA (µg/kg)	Mean dietary exposure	p95 Dietary exposure	Mean PAH4 (µg/kg)	Mean dietary exposurea	p95 Dietary exposure	Mean AA (µg/kg)	Mean dietary exposure	p95 Dietary exposure
		LBb	UBc	LB	UB	LB	UB	LBc			UBd			LB
Cereals (n = 22)	1–2	0.12	0.43	1.51	5.42	2.77	9.93	16.52	208.30	381.22	0.37	4.69	8.59	0.57	7.69	13.10
	3–5			1.31	4.69	2.38	8.52		180.28	327.36		4.06	7.38		6.72	11.25
	6–11			0.96	3.44	1.69	6.07		131.98	233.20		2.97	5.26		4.39	8.01
	12–18			0.63	2.26	1.24	4.45		86.86	170.83		1.96	3.85		2.81	5.87
	19–29			0.51	1.83	1.09	3.91		70.12	150.03		1.58	3.38		2.37	5.15
	30–49			0.50	1.79	1.07	3.85		68.85	147.84		1.55	3.33		2.25	5.08
	50–64			0.50	1.80	1.02	3.67		69.29	140.79		1.56	3.17		2.36	4.84
	Over 65			0.48	1.73	0.91	3.25		66.37	124.69		1.50	2.81		2.24	4.28
Vegetables (n = 24)	1–2	0.07	0.36	0.36	1.85	1.13	5.79	501.74	2579.11	8074.02	0.33	1.71	5.36	0.56	3.26	8.99
	3–5			0.35	1.78	0.88	4.55		2475.08	6335.19		1.64	4.21		2.94	7.05
	6–11			0.30	1.55	0.76	3.90		2167.08	5434.65		1.44	3.61		2.24	6.05
	12–18			0.24	1.23	0.58	2.97		1713.58	4140.06		1.14	2.75		1.62	4.61
	19–29			0.22	1.15	0.58	2.99		1607.61	4166.16		1.07	2.77		1.63	4.64
	30–49			0.31	1.58	0.72	3.69		2204.48	5137.69		1.46	3.41		2.31	5.72
	50–64			0.38	1.95	0.82	4.23		2715.88	5901.92		1.80	3.92		2.84	6.57
	Over 65			0.35	1.82	0.87	4.45		2540.95	6205.41		1.69	4.12		2.88	6.91
Root & tuber crops (n = 10)	1–2	0.02	0.34	0.04	0.66	0.16	2.71	583.89	1135.24	4650.12	0.38	0.74	3.02	0.56	1.16	4.47
	3–5			0.02	0.41	0.11	1.90		708.37	3261.47		0.46	2.12		0.70	3.14
	6–11			0.02	0.28	0.07	1.26		488.00	2155.59		0.32	1.40		0.42	2.07
	12–18			0.01	0.18	0.05	0.84		309.02	1435.91		0.20	0.93		0.24	1.38
	19–29			0.01	0.18	0.05	0.86		312.59	1484.49		0.20	0.96		0.29	1.43
	30–49			0.01	0.15	0.04	0.70		264.31	1198.51		0.17	0.78		0.23	1.15
	50–64			0.01	0.20	0.06	1.10		349.47	1885.87		0.23	1.23		0.30	1.81
	Over 65			0.01	0.21	0.07	1.20		364.15	2064.70		0.24	1.34		0.40	1.99
Mushrooms (n = 8)	1–2	0.00	0.33	0.00	0.07	0.00	0.28	15.43	3.33	13.17	1.29	0.28	1.10	2.11	0.57	1.80
	3–5			0.00	0.07	0.00	0.37		3.32	17.19		0.28	1.43		0.45	2.35
	6–11			0.00	0.05	0.00	0.19		2.23	8.91		0.19	0.74		0.22	1.22
	12–18			0.00	0.03	0.00	0.15		1.36	6.92		0.11	0.58		0.14	0.95
	19–29			0.00	0.03	0.00	0.14		1.63	6.47		0.14	0.54		0.27	0.88
	30–49			0.00	0.04	0.00	0.22		1.95	10.42		0.16	0.87		0.22	1.42
	50–64			0.00	0.03	0.00	0.16		1.45	7.48		0.12	0.62		0.23	1.02
	Over 65			0.00	0.03	0.00	0.11		1.36	5.37		0.11	0.45		0.13	0.73
Nuts & seeds (n = 4)	1–2	0.03	0.34	0.00	0.04	0.02	0.17	359.52	41.59	182.87	0.36	0.04	0.18	0.50	0.14	0.25
	3–5			0.00	0.04	0.02	0.19		38.32	201.49		0.04	0.20		0.04	0.28
	6–11			0.00	0.03	0.01	0.11		29.17	117.33		0.03	0.12		0.02	0.16
	12–18			0.00	0.01	0.00	0.04		15.03	39.78		0.01	0.04		0.03	0.06
	19–29			0.00	0.02	0.01	0.06		18.85	64.91		0.02	0.06		0.02	0.09
	30–49			0.00	0.03	0.01	0.11		28.37	113.17		0.03	0.11		0.03	0.16
	50–64			0.00	0.05	0.01	0.17		47.64	176.01		0.05	0.17		0.07	0.24
	Over 65			0.00	0.04	0.02	0.18		41.10	188.19		0.04	0.19		0.07	0.26
Fruits (n = 1)	1–2	0.00	0.33	0.00	3.59	0.00	12.26	3.07	33.43	114.10	0.32	3.52	12.00	0.53	3.91	19.82
	3–5			0.00	2.67	0.00	7.86		24.90	73.20		2.62	7.70		4.03	12.71
	6–11			0.00	1.32	0.00	4.85		12.27	45.17		1.29	4.75		1.89	7.84
	12–18			0.00	0.48	0.00	2.30		4.51	21.40		0.47	2.25		0.77	3.72
	19–29			0.00	0.43	0.00	2.04		3.99	18.99		0.42	2.00		0.52	3.30
	30–49			0.00	0.68	0.00	2.70		6.37	25.12		0.67	2.64		0.98	4.36
	50–64			0.00	1.01	0.00	3.49		9.43	32.50		0.99	3.42		1.40	5.64
	Over 65			0.00	0.89	0.00	3.17		8.28	29.49		0.87	3.10		1.44	5.12
Fish & shellfish (n = 41)	1–2	0.38	0.64	1.02	1.73	4.16	7.01	92.31	12.91	52.47	0.66	1.77	7.20	1.08	1.02	11.86
	3–5			0.98	1.64	3.37	5.68		12.29	42.47		1.69	5.83		1.27	9.60
	6–11			0.64	1.08	2.30	3.88		8.08	29.03		1.11	3.98		0.87	6.56
	12–18			0.51	0.86	2.06	3.48		6.41	26.00		0.88	3.57		0.52	5.88
	19–29			0.46	0.78	2.00	3.38		5.80	25.26		0.80	3.46		0.53	5.71
	30–49			0.60	1.00	2.20	3.71		7.51	27.73		1.03	3.80		0.82	6.27
	50–64			0.71	1.19	2.64	4.45		8.90	33.26		1.22	4.56		0.82	7.52
	Over 65			0.67	1.14	2.89	4.87		8.49	36.46		1.16	5.00		0.57	8.24
Sea algae (n = 7)	1–2	0.56	0.78	0.60	0.84	2.62	3.65	92.31	98.88	432.08	1.66	1.78	7.78	4.01	0.90	18.79
	3–5			0.62	0.87	2.54	3.54		102.66	419.35		1.85	7.55		0.40	18.23
	6–11			0.34	0.48	1.43	2.00		56.24	236.32		1.01	4.26		0.26	10.28
	12–18			0.21	0.29	0.94	1.30		34.87	154.32		0.63	2.78		0.10	6.71
	19–29			0.15	0.22	0.63	0.88		25.52	103.94		0.46	1.87		0.07	4.52
	30–49			0.21	0.29	1.10	1.53		34.35	180.89		0.62	3.26		0.19	7.87
	50–64			0.28	0.39	1.51	2.11		45.99	249.57		0.83	4.50		0.25	10.85
	Over 65			0.29	0.40	1.86	2.59		47.15	306.78		0.85	5.53		0.18	13.34
Meat (n = 66)	1–2	0.08	0.39	0.29	1.41	0.84	4.08	7.32	26.40	76.56	0.49	1.78	5.15	0.86	3.66	9.02
	3–5			0.33	1.60	0.95	4.62		30.05	86.70		2.02	5.83		3.53	10.21
	6–11			0.28	1.34	0.77	3.77		25.23	70.70		1.70	4.76		2.74	8.32
	12–18			0.24	1.15	0.66	3.24		21.61	60.81		1.45	4.09		2.01	7.16
	19–29			0.24	1.15	0.74	3.63		21.58	68.05		1.45	4.58		1.94	8.01
	30–49			0.18	0.86	0.56	2.72		16.14	51.08		1.09	3.44		1.88	6.01
	50–64			0.12	0.58	0.44	2.16		10.81	40.57		0.73	2.73		1.23	4.78
	Over 65			0.08	0.39	0.36	1.74		7.38	32.72		0.50	2.20		0.68	3.85
Eggs (n = 1)	1–2	0.00	0.33	0.00	0.70	0.00	2.31	0.32	0.67	2.21	0.76	1.61	5.32	1.19	1.95	8.36
	3–5			0.00	0.61	0.00	2.02		0.58	1.93		1.41	4.65		2.08	7.30
	6–11			0.00	0.34	0.00	1.16		0.33	1.11		0.79	2.68		1.23	4.21
	12–18			0.00	0.18	0.00	0.73		0.17	0.70		0.42	1.67		0.72	2.63
	19–29			0.00	0.17	0.00	0.67		0.17	0.64		0.40	1.54		0.59	2.42
	30–49			0.00	0.18	0.00	0.64		0.17	0.61		0.42	1.47		0.63	2.31
	50–64			0.00	0.18	0.00	0.69		0.17	0.66		0.42	1.59		0.62	2.50
	Over 65			0.00	0.13	0.00	0.61		0.13	0.58		0.31	1.41		0.48	2.21

^aTC_PAH4 calculated by the sum of the concentrations of four PAH compounds

^bMean and P95 dietary exposure unit: ng/kg bw/day

^cLower bound (LB) applied zero for the value of below LOD

^dUpper bound (UB) applied LOD for the value of below LOD

^e1/2 LOD applied half of LOD for the value of below LOD

Table 4.

The margin of exposure (MOE) of PAH4, AA, and PhIP

Categories	Age group (years)	PAH4a				AAb		PhIPc
		MOE (Mean dietary exposure)d		MOE (P95 Dietary exposure)		MOE (Mean dietary exposure)	MOE (P95 Dietary exposure)	MOE (Mean dietary exposure)	MOE (P95 Dietary exposure)
		LB5)	UBf	LB	UB	1/2 LODg	1/2 LOD	1/2 LOD	1/2 LOD
Cereals (n = 22)	1–2	2.24 × 105	6.26 × 104	1.22 × 105	3.42 × 104	8.16 × 102	4.46 × 102	1.57 × 105	8.61 × 104
	3–5	2.59 × 105	7.24 × 104	1.42 × 105	3.98 × 104	9.43 × 102	5.19 × 102	1.82 × 105	1.00 × 105
	6–11	3.54 × 105	9.89 × 104	2.00 × 105	5.59 × 104	1.28 × 103	7.29 × 102	2.48 × 105	1.40 × 105
	12–18	5.38 × 105	1.50 × 105	2.73 × 105	7.64 × 104	1.95 × 103	9.95 × 102	3.78 × 105	1.92 × 105
	19–29	6.67 × 105	1.86 × 105	3.11 × 105	8.70 × 104	2.42 × 103	1.13 × 103	4.68 × 105	2.18 × 105
	30–49	6.79 × 105	1.89 × 105	3.16 × 105	8.83 × 104	2.46 × 103	1.15 × 103	4.76 × 105	2.22 × 105
	50–64	6.75 × 105	1.88 × 105	3.32 × 105	9.27 × 104	2.45 × 103	1.20 × 103	4.73 × 105	2.33 × 105
	Over 65	7.05 × 105	1.96 × 105	3.75 × 105	1.04 × 105	2.56 × 103	1.36 × 103	4.94 × 105	2.63 × 105
Vegetables (n = 24)	1–2	9.44 × 105	1.83 × 105	3.01 × 105	5.86 × 104	6.60 × 101	0.21 × 101	4.32 × 105	1.38 × 105
	3–5	9.84 × 105	1.91 × 105	3.84 × 105	7.47 × 104	6.90 × 101	0.27 × 101	4.50 × 105	1.75 × 105
	6–11	1.12 × 106	2.18 × 105	4.46 × 105	8.71 × 104	7.80 × 101	0.31 × 101	5.14 × 105	2.05 × 105
	12–18	1.42 × 106	2.76 × 105	5.88 × 105	1.14 × 105	9.90 × 101	0.41 × 101	6.50 × 105	2.69 × 105
	19–29	1.51 × 106	2.94 × 105	5.84 × 105	1.13 × 105	1.06 × 102	0.41 × 101	6.93 × 105	2.67 × 105
	30–49	1.10 × 106	2.14 × 105	4.74 × 105	9.22 × 104	7.70 × 101	0.33 × 101	5.05 × 105	2.16 × 105
	50–64	8.97 × 105	1.74 × 105	4.12 × 105	8.02 × 104	6.30 × 101	0.29 × 101	4.10 × 105	1.88 × 105
	Over 65	9.59 × 105	1.86 × 105	3.92 × 105	7.63 × 104	6.70 × 101	0.27 × 101	4.38 × 105	1.79 × 105
Root & tuber crops (n = 10)	1–2	8.74 × 106	5.14 × 105	2.13 × 106	1.25 × 105	1.50 × 102	0.37 × 101	1.00 × 106	2.44 × 105
	3–5	1.40 × 107	8.24 × 105	3.04 × 106	1.79 × 105	2.40 × 102	0.52 × 101	1.60 × 106	3.49 × 105
	6–11	2.03 × 107	1.19 × 106	4.60 × 106	2.70 × 105	3.48 × 102	0.79 × 101	2.33 × 106	5.28 × 105
	12–18	3.21 × 107	1.88 × 106	6.91 × 106	4.06 × 105	5.50 × 102	1.18 × 102	3.68 × 106	7.93 × 105
	19–29	3.17 × 107	1.86 × 106	6.68 × 106	3.93 × 105	5.44 × 102	1.15 × 102	3.64 × 106	7.67 × 105
	30–49	3.75 × 107	2.20 × 106	8.28 × 106	4.87 × 105	6.43 × 102	1.42 × 102	4.30 × 106	9.50 × 105
	50–64	2.84 × 107	1.67 × 106	5.26 × 106	3.09 × 105	4.86 × 102	0.90 × 101	3.25 × 106	6.03 × 105
	Over 65	2.72 × 107	1.60 × 106	4.80 × 106	2.82 × 105	4.67 × 102	0.82 × 101	3.12 × 106	5.51 × 105
Mushrooms (n = 8)	1–2	–	4.78 × 106	–	1.20 × 106	5.11 × 104	1.29 × 104	2.66 × 106	6.73 × 105
	3–5	–	4.79 × 106	–	9.25 × 105	5.12 × 104	9.88 × 103	2.67 × 106	5.16 × 105
	6–11	–	7.13 × 106	–	1.78 × 106	7.62 × 104	1.90 × 104	3.97 × 106	9.95 × 105
	12–18	–	1.16 × 107	–	2.29 × 106	1.24 × 105	2.45 × 104	6.52 × 106	1.28 × 106
	19–29	–	9.76 × 106	–	2.45 × 106	1.04 × 105	2.62 × 104	5.45 × 106	1.37 × 106
	30–49	–	8.16 × 106	–	1.52 × 106	8.73 × 104	1.63 × 104	4.55 × 106	8.51 × 105
	50–64	–	1.10 × 107	–	2.12 × 106	1.17 × 105	2.27 × 104	6.14 × 106	1.18 × 106
	Over 65	–	1.17 × 107	–	2.96 × 106	1.25 × 105	3.16 × 104	6.54 × 106	1.65 × 106
Nuts & seeds (n = 4)	1–2	9.79 × 107	8.64 × 106	2.28 × 107	1.96 × 106	4.08 × 103	9.30 × 102	1.79 × 107	4.07 × 106
	3–5	1.06 × 108	9.38 × 106	2.02 × 107	1.78 × 106	4.43 × 103	8.44 × 102	1.94 × 107	3.69 × 106
	6–11	1.39 × 108	1.23 × 107	3.47 × 107	3.06 × 106	5.82 × 103	1.44 × 103	2.55 × 107	6.34 × 106
	12–18	2.71 × 108	2.39 × 107	1.02 × 108	9.03 × 106	1.13 × 104	4.27 × 103	4.95 × 107	1.87 × 107
	19–29	2.16 × 108	1.90 × 107	6.27 × 107	5.53 × 106	9.01 × 103	2.61 × 103	3.94 × 107	1.14 × 107
	30–49	1.43 × 108	1.26 × 107	3.60 × 107	3.17 × 106	5.99 × 103	1.50 × 103	2.62 × 107	6.57 × 106
	50–64	8.55 × 107	7.54 × 106	2.31 × 107	2.04 × 106	3.56 × 103	9.66 × 102	1.56 × 107	4.22 × 106
	Over 65	9.91 × 107	8.74 × 106	2.16 × 107	1.91 × 106	4.13 × 103	9.03 × 102	1.81 × 107	3.95 × 106
Fruits (n = 1)	1–2	–	9.46 × 104	–	2.77 × 104	5.08 × 103	1.49 × 103	2.10 × 105	6.16 × 104
	3–5	–	1.27 × 105	–	4.32 × 104	6.82 × 103	2.32 × 103	2.82 × 105	9.61 × 104
	6–11	–	2.57 × 105	–	7.00 × 104	1.38 × 104	3.76 × 103	5.73 × 105	1.55 × 105
	12–18	–	7.01 × 105	–	1.47 × 105	3.77 × 104	7.94 × 103	1.56 × 106	3.28 × 105
	19–29	–	7.92 × 105	–	1.66 × 105	4.25 × 104	8.95 × 103	1.76 × 106	3.70 × 105
	30–49	–	4.96 × 105	–	1.25 × 105	2.66 × 104	6.76 × 103	1.10 × 106	2.80 × 105
	50–64	–	3.35 × 105	–	9.73 × 104	1.80 × 104	5.23 × 103	7.46 × 105	2.16 × 105
	Over 65	–	3.82 × 105	–	1.07 × 105	2.05 × 104	5.76 × 103	8.50 × 105	2.38 × 105
Fish & shellfish (n = 41)	1–2	3.31 × 105	1.96 × 105	8.16 × 104	4.84 × 104	1.31 × 104	3.24 × 103	4.17 × 105	1.02 × 105
	3–5	3.48 × 105	2.06 × 105	1.00 × 105	5.98 × 104	1.38 × 104	4.00 × 103	4.39 × 105	1.27 × 105
	6–11	5.30 × 105	3.14 × 105	1.47 × 105	8.76 × 104	2.10 × 104	5.85 × 103	6.67 × 105	1.85 × 105
	12–18	6.68 × 105	3.96 × 105	1.64 × 105	9.78 × 104	2.65 × 104	6.53 × 103	8.41 × 105	2.07 × 105
	19–29	7.38 × 105	4.38 × 105	1.69 × 105	1.00 × 105	2.93 × 104	6.73 × 103	9.30 × 105	2.13 × 105
	30–49	5.70 × 105	3.38 × 105	1.54 × 105	9.17 × 104	2.26 × 104	6.13 × 103	7.18 × 105	1.94 × 105
	50–64	4.81 × 105	2.85 × 105	1.28 × 105	7.64 × 104	1.90 × 104	5.11 × 103	6.06 × 105	1.62 × 105
	Over 65	5.04 × 105	2.99 × 105	1.17 × 105	6.97 × 104	2.00 × 104	4.66 × 103	6.35 × 105	1.47 × 105
Sea algae (n = 7)	1–2	5.66 × 105	4.06 × 105	1.29 × 105	9.31 × 104	1.71 × 103	3.93 × 102	4.15 × 105	9.50 × 104
	3–5	5.45 × 105	3.91 × 105	1.33 × 105	9.59 × 104	1.65 × 103	4.05 × 102	4.00 × 105	9.79 × 104
	6–11	9.96 × 105	7.15 × 105	2.37 × 105	1.70 × 105	3.02 × 103	7.19 × 102	7.30 × 105	1.73 × 105
	12–18	1.60 × 106	1.15 × 106	3.63 × 105	2.60 × 105	4.87 × 103	1.10 × 103	1.17 × 106	2.66 × 105
	19–29	2.19 × 106	1.57 × 106	5.39 × 105	3.87 × 105	6.66 × 103	1.63 × 103	1.60 × 106	3.95 × 105
	30–49	1.63 × 106	1.17 × 106	3.09 × 105	2.22 × 105	4.94 × 103	9.40 × 102	1.19 × 106	2.27 × 105
	50–64	1.21 × 106	18.75 × 105	2.24 × 105	1.61 × 105	3.69 × 103	6.81 × 102	8.93 × 105	1.64 × 105
	Over 65	1.18 × 106	8.53 × 105	1.82 × 105	1.31 × 105	3.60 × 103	5.54 × 102	8.71 × 105	1.33 × 105
Meat (n = 66)	1–2	4.00 × 106	2.41 × 105	4.06 × 105	8.33 × 104	6.44 × 103	2.22 × 103	4.16 × 105	1.43 × 105
	3–5	3.51 × 106	2.12 × 105	3.58 × 105	7.36 × 104	5.65 × 103	1.96 × 103	3.65 × 105	1.26 × 105
	6–11	4.18 × 106	2.52 × 105	4.40 × 105	9.02 × 104	6.73 × 103	2.40 × 103	4.35 × 105	1.55 × 105
	12–18	4.89 × 106	2.95 × 105	5.11 × 105	1.04 × 105	7.86 × 103	2.79 × 103	5.08 × 105	1.80 × 105
	19–29	4.89 × 106	2.95 × 105	4.57 × 105	9.37 × 104	7.87 × 103	2.49 × 103	5.09 × 105	1.61 × 105
	30–49	6.54 × 106	3.95 × 105	6.09 × 105	1.24 × 105	1.05 × 104	3.32 × 103	6.81 × 105	2.15 × 105
	50–64	9.77 × 106	5.90 × 105	7.66 × 105	1.57 × 105	1.57 × 104	4.19 × 103	1.01 × 106	2.71 × 105
	Over 65	1.43 × 106	8.64 × 105	9.50 × 105	1.95 × 105	2.30 × 104	5.19 × 103	1.48 × 106	3.36 × 105
Eggs (n = 1)	1–2	–	4.86 × 105	–	1.46 × 105	2.54 × 105	7.69 × 104	4.60 × 105	1.39 × 105
	3–5	–	5.55 × 105	–	1.68 × 105	2.91 × 105	8.81 × 104	5.26 × 105	1.59 × 105
	6–11	–	9.86 × 105	–	2.91 × 105	5.16 × 105	1.52 × 105	9.34 × 105	2.76 × 105
	12–18	–	1.86 × 106	–	4.66 × 105	9.78 × 105	2.44 × 105	1.76 × 106	4.41 × 105
	19–29	–	1.94 × 106	–	5.07 × 105	1.01 × 106	2.65 × 105	1.84 × 106	4.80 × 105
	30–49	–	1.85 × 106	–	5.31 × 105	9.71 × 105	2.78 × 105	1.75 × 106	5.03 × 105
	50–64	–	1.85 × 106	–	4.90 × 105	9.72 × 105	2.57 × 105	1.75 × 106	4.64 × 105
	Over 65	–	2.54 × 106	–	5.56 × 105	1.33 × 106	2.91 × 105	2.40 × 106	5.26 × 105

Bold formatted numbers indicate MOEs below 10,000, which are considered potentially concerning

^aBMDL₁₀ for PAH4: 0.34 mg/kg bw/d

^bBMDL₁₀ for AA: 0.17 mg/kg bw/d

^cBMDL₁₀ for PhIP:0.74 mg/kg bw/d

^dUnder the 10,000 value was considered “may be a concern”

^eLower bound (LB): Zero applied for the values of below LOD

^fUpper bound (UB): LOD applied for the values of below LOD

^g1/2 LOD: Half of LOD applied for the values of below LOD

Mean dietary exposure was calculated using daily food intake rate ($\text{IRi}$) and body weight ($\text{bw}$) data from Korea Health Industry Development Institute, conducted in 2019. The equations below were employed to calculate mean and 95% dietary exposures for PAH4, HAA8, PhIP, and AA:

$$\text{Dietary exposure of PAH}4=\sum _{i=1}^n\frac{\text{TCPAH4i}\times \text{IRi}}{\text{bw}}\left(\frac{\text{ng}}{\text{kg b}.\text{w}.\text{day}}\right)$$ 2

$$\text{Dietary exposure of HAA}8,\text{PhIP},\text{and AA}=\sum _{i=1}^n\frac{\text{Ci}\times \text{IRi}}{\text{bw}}\left(\frac{\text{ng}}{\text{kg b}.\text{w}.\text{day}}\right)$$ 3

The MOE approach for risk assessment involved calculating the ratio of the bench marking response of 10% (BMLD₁₀) to daily exposure (EFSA, 2022) using Eq. 4:

$$\text{MOE}=\frac{\text{BMDL}10}{\text{Daily exposure}}$$ 4

The BMDL₁₀ of PAH4 was used at 0.34 mg/kg bw/d (EFSA, 2008), while 0.17 mg/kg bw/d for AA was used (EFSA, 2015), and 0.74 mg/kg bw/d for PhIP (Carthew et al., 2010). The MOE of PhIP was estimated only among HAAs. MOE values over 10,000 were considered to be “low concern” regarding carcinogenicity, while those below 10,000 were considered to “may be a concern” (COC, 2022).

Statistical analysis

Statistical analyses were performed using t-test by SAS version 9.4 (SAS Inc., Cary, NC, USA), with significance levels set at *P < 0.05 and **P < 0.01.

Results and discussion

PAHs, AA, and HAAs concentrations in foods prepared in an air fryer

Table 2 presents the concentrations of PAHs, AA, and HAAs in various food categories cooked with an air fryer. The occurrence of PAH4, AA, HAA8, and HAA10 in various food categories cooked using the same cooking method is shown in Fig. 1.

Fig. 1.

The tendency of occurrence for the polycyclic aromatic hydrocarbons (PAHs), acrylamide (AA), and heterocyclic aromatic amines (HAAs) by food categories. A The sum of four PAHs (PAH4) including benzo[a]pyrene, benz[a]anthracene, benzo[b]fluoranthene, and chrysene. B The tendency of occurrence for AA. C The sum of eight HAAs (HAA8) including 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), 2-amino-3,4-dimethyl-3H-imidazo[4,5-f] quinoline (MeIQ), 2-amino-3,8-dimethylimidazo[4,5-f]quinoline] quinoxaline (MeIQx), 2-amino-1-methyl-6-phenylimidazo[4,5-b] pyridine (PhIP), 2-amino-9H-pyrido[2,3-b]indole (AαC), 2-amino-3-methyl-9H-pyrido[2,3-b] indole (MeAαC), 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole acetate (Trp-P-1), and 3-amino-1-methyl-5H-pyrido[4,3-b]indole acetate (Trp-P-2). D The sum of ten HAAs (HAA10) including HAA8, 1-methyl-9H-pyrido[3,4-b]indole (harman), and 9H-pyrido[3,4-b]indole (norharman). The thick line means the median value in boxplots, the o marker means an outlier, and the + marker indicates a mean value

In PAHs, Chr was the most frequently detected compound, found in 36 samples. This aligns with the EFSA (2008) report stating that Chr is the most common PAH in food among PAH4. For PAH4, the concentration in all the food samples ranged from 0 to 5.20 µg/kg, with the highest value detected in fish & shellfish (Table 2). The highest PAH4 mean was in sea algae (0.56 µg/kg), followed by fish & shellfish (0.38 µg/kg). Fish & shellfish and sea algae groups had relatively higher values than other groups in each PAH. Environmental factors, such as water pollution, can influence the presence of PAHs, suggesting a possible exposure of fishery products to these compounds (Phillips, 1999). The cooking method, including smoking, grilling, and frying, can also impact the origin of PAHs in food (Zelinkova and Wenzl, 2015). In a Chinese study, PAH4 levels in processed fish were quantified at 5.02 and 2.54 µg/kg in grilling and frying, respectively (Wang et al., 2021). Although fish & shellfish showed relatively high PAH4 levels (mean = 0.38 µg/kg) in this study, they were about 6- and 13-fold lower than frying (2.54 µg/kg) and grilling (5.02 µg/kg) data from Wang et al. (2021), respectively. The use of an air fryer, which employs high-temperature air convection without a direct flame, has the potential to decrease the formation of PAHs by avoiding flames during the cooking process (Chen and Lin, 1997). This outcome implies that air fryers may be considered effective in diminishing PAH4 levels.

The concentrations of AA in food cooked with the air fryer, are shown in Table 2. The highest mean value of AA was detected in root & tuber crops (583.89 µg/kg), followed by vegetable (501.74 µg/kg) and nut & seed (359.52 µg/kg). In AA, all sample categories were measured with over a 95% detection rate, except for eggs. The contents of AA exhibit variations according to temperature, carbohydrate, amino acid, and food category. Previous research on deep-fried processed foods showed AA content ranging from 150 to 1,000 µg/kg in carbohydrate-rich foods (potato and beetroot) and 5–50 µg/kg in protein-rich foods (beef, chicken, and fish) (Tareke et al., 2002). In comparison, this study indicated a wide range in carbohydrate-rich foods (root & tuber crops: 1.07–2645.76 µg/kg) and a similar range in protein-rich foods (meat: 0.66–38.65 µg/kg). The formation of AA is triggered by high temperature, prolonged cooking time, the presence of reducing sugar, relatively low water content, and water activity (Anese et al., 2009). The cooking principle of air fryers, which employs hot air convection and reduces the moisture content of food materials, can lead to higher AA formation compared to other cooking methods. Consequently, air fryers may play a significant role in the generation of AA.

The concentrations of HAAs in food cooked using an air fryer, are categorized by food type in Table 2. For HAA8, the highest mean value was in sea algae (3.91 µg/kg) followed by mushrooms (2.00 µg/kg), while the lowest value was nut & seed (0.34 µg/kg). The dry condition of mushrooms is composed of abundant carbohydrates, including glucans, mono- and disaccharides, sugar alcohols, and glycan which can react with the Maillard reaction (Kurtzman Jr, 1997). Therefore, it can be assumed that the composition or moisture content of mushroom and sea algae can affect the formation of HAAs. In the case of HAA10, cereal displayed the highest mean concentration (46.50 µg/kg), followed by meat (46.35 µg/kg) and fish & shellfish (42.04 µg/kg), with fruit displaying the lowest value (2.98 µg/kg). When it comes to the detection rate of HAAs, IQ, which is the most carcinogenic compound among HAAs, was not detected in any of the samples. On the other hand, norharman showed a 100% detection rate in all sampled categories, and harman had a 100% detection rate in all sample categories except for vegetables, and root & tuber crops. In this study, PhIP, which is frequently generated during the cooking process, demonstrated a high detection rate exceeding 40% in all sample categories, in addition, norharman and harman had the most frequent detection rate (Table 2). The fish & shellfish, and meat groups are ranked fourth and fifth in this study concerning HAA8. Despite meat and fish containing high fat, the formation of HAA can be reduced due to the partial loss of their precursors and amino acids during the cooking process (Neves et al., 2021).

In comparison to previous research, the air fryer may be a suitable factor for reducing PAHs and HAAs compared to conventional cooking methods. However, the level of AA is relatively higher in carbohydrate-rich foods than in conventional cooking methods and is notably high in non-carbohydrate-rich food categories such as sea algae. Although all three hazardous substances are products of the Maillard reaction, the detection rate or concentration of these compounds varies significantly with different cooking methods.

Correlation analysis of three hazardous substances in agricultural products, fishery products, and livestock products

In a study conducted by EFSA (2008), it was concluded that B[a]P alone is not a suitable indicator to estimate carcinogenic and genotoxic PAHs in food. They suggested that PAH4 is a more appropriate indicator for risk assessment of PAHs. Additionally, the IARC identified certain compounds, including IQ, MeIQ, MelQx, PhIP, AαC, MeAαC, Trp-1, and Trp-2 as direct mutagenic compounds in HAAs. However, harman and norharman were not classified as carcinogenic compounds (IARC, 2023). Thus, in the analysis of carcinogenic and genotoxicity indicators using boxplots, correlation analysis, and estimated dietary exposure, PAH4 and HAA8 were chosen. The correlation among three hazardous substances − PAH4, AA, and HAA8 − was investigated by Spearman correlation in agricultural, livestock, and fishery products based on Table 2 (Fig. 2).

Fig. 2.

Correlation analysis of three hazardous substances by food categories. Spearman correlation of PAH4, AA, and HAA8 in agricultural products (A), fishery products (B), and livestock products (C). The color gradient from red to blue indicates positive and negative correlations. **P < 0.01 and *P < 0.05 are considered as statistically significant

In agricultural products, positive correlations were observed between PAH4 and AA, AA and HAA8, and PAH4 and HAA8, although none were statistically significant (Fig. 2A). In livestock products, a significantly (*P < 0.05) positive correlation was found between PAH4 and AA (Fig. 2C). A significant (**P < 0.01) negative correlation was observed between PAH4 and HAA8, while the correlation between AA and HAA8 was negative but negligible. Fishery products showed significant (**P < 0.01) positive correlations between PAH4 and HAA8, and a positive correlation between PAH4 and AA, although not statistically significant (Fig. 2B).

Generally, PAHs, AA, and HAAs are easily generated by high temperatures and long cooking times (Adeyeye, 2020; Skog et al., 1998; Zhang et al., 2009). Thus, Fig. 2 mostly shows positive correlations in samples. However, a significant negative correlation between PAH4 and HAA8 was observed in livestock products. Previous research indicated a positive relationship between fat content and PAHs (B[a]P and B[a]A) and a negative relationship between fat content and HAAs (IQ, MeIQ,4,8-DiMeIQx, and PhIP), represented by a positive (r = 0.414, P < 0.05) and negative (r =  − 0.213, respectively) correlation (Wang et al., 2019). Lipid pyrolysis generally generates the formation of PAHs, but an excessive amount of fat inhibits the production of HAAs as its precursors and amino acids are partially lost (Neves et al., 2021). The addition of marinade in meat may have affected the negative correlation between PAHs and HAAs formation. Nor Hasyimah et al. (2022) examined the effect of marinade on the production of these two compounds in beef. In grilling methods, marinade (salt, oil, honey, cumin, fennel, coriander, turmeric, galangal, and lemon grass) increased the formation of PAHs up to 43.51%. However, the formation of HAAs decreased up to 84.54%.

Additionally, the correlation between PAH4 and AA shows a positive relationship in livestock products (Fig. 2C). Although the correlation between PAHs and AA in meat has been extensively studied, the levels of PAH4 and AA in chicken cooked with an air fryer exhibited a similar increasing trend in breast, thighs, and wings after the thawing method (Lee et al., 2020). These results suggest that using an air fryer in meat shows a positive correlation between PAH4 and AA (Fig. 2B). Notably, this study shows a significant difference between PAH4 and HAA8, and AA and HAA8 in fishery products. A similar trend was also observed in previous studies. Fat content and PAH8 had a positive correlation in fish (trout, sea bream, seabass, salmon, shad) (r = 0.506, P < 0.01), and a positive correlation existed between PAH4 and HAA8 when the fish was grilled (r = 0.942, P < 0.01) (Oz, 2021). Seasoning can affect the positive correlation between PAH4 and AA, as many of seasonings were used for cooking in this study. PAH4 was detected in herbs and spices (2.86–25.76 µg/kg) (Rozentale et al., 2018) and AA was detected in seasoning, including black pepper, sauces, curries, stock powders, and spices (397, 8.6, 70.6, 14.0, and 124.4 µg/kg, respectively) (Jeong et al., 2020).

Considering these results, fat content, marinade, and seasoning have an impact on the production of PAH4, AA, and HAA8; therefore, the correlation of three hazardous substances is expected to be affected by factors such as food composition, cooking methods, marinade, and seasoning in this study.

Mean and P95 dietary exposure and MOE in the Korean population

Tables 3 and 4 present the mean and P95 dietary exposure of PAH4, AA, PhIP, and HAA8, along with the corresponding MOE calculations for risk assessment. The highest mean concentration of AA was detected in root & tuber crops. However, the mean and P95 dietary exposure of root & tuber crops were lower than those of vegetables in all age groups due to a lower intake of root & tuber crops compared to vegetables in the Korean population. For PhIP and HAA8, sea algae exhibited the highest mean value (1.66 and 4.01 µg/kg, respectively) (Table 3). Nevertheless, in mean dietary exposure, cereals and vegetables were the highest groups in the 1–49 years age range and over 50 years, respectively. This suggests a higher mean food intake rate for cereals and vegetables than sea algae in the Korean population. In PAH4, cereal had the highest mean dietary exposure in all age groups. Notably, fish and shellfish showed a high dietary exposure to PAH4 in children under 5 years.

According to a study in China, the mean dietary exposure of AA, based on daily intake rates by food categories in 2007, measured 0.076, 0.136, 0.023, 0.016, 0.001, 0.006, and 0.019 µg/kg bw/d in cereal, vegetable, potato (root & tuber crops), legumes and nut (nuts & seeds), fruit, aquatic foods (fish & shellfish), and meat, respectively (Zhou et al., 2013). The results of these categories in this study showed much higher dietary exposure than China. Additionally, in children, the dietary exposure of AA was 1 µg/kg bw/d in the mean daily intake and 4 µg/kg bw/d in high-risk intake and it has been reported to be double compared to adults (WHO, 2011). This study shows the same tendency, with higher dietary exposure in children, and the dietary exposure of AA is relatively higher than in other countries.

In HAA8, cereal and vegetable were the highest values of mean dietary exposure in the 1–29 years age group and over 30 years age group, respectively. Cereal was the highest value in the 30–49 years, 50–64 years, and over 65 years age groups (2.25, 2.36, and 2.24 µg/kg bw/d, respectively). Although there are few studies on dietary exposure to HAA8, some studies have been conducted on PhIP in meat. In the USA, dietary exposure to PhIP ranged from 7.7 to 10.8 ng/kg bw/d in children and from 9.2 to 12.7 ng/kg bw/d in adults (Keating and Bogen, 2004). Among HAA8, PhIP is known to exist most abundantly in meat. However, our result shows a lower dietary exposure compared to other countries since the main food intake in Asia is grains, not meat.

According to EFSA, the MOE approach is the most appropriate risk assessment in carcinogenic and genotoxic compounds such as PAHs, AA, and HAAs. Also, a value lower than 10,000 for MOE is considered a concern (COC, 2022). Values lower than 10,000 are shown in bold letters (Table 4). In P95 dietary exposure, the MOE value of AA may be a concern in all age groups except for mushrooms and eggs. In mean dietary exposure, the MOE value of AA may be a concern in cereals, vegetables, root & tuber crops, and sea algae in all age groups. Also, nuts & seeds may be a concern in all age groups except for 12–18 years, and fruits and meat were in the “may be a concern” range in 1–5 years and 1–29 years, respectively. However, the number of samples in fruits is one. Thus, it is difficult to say that all types of fruits represent the carcinogenic concern range. None of the values were within the “may be a concern” range in PAH4 and PhIP.

Based on MOE values, when using the air fryer, it is essential to consider the intake of cereals, vegetables, root & tuber crops, nuts & seeds, fruits, sea algae, and meat in the young population. High-intake consumers need to be cautious when consuming cereals, vegetables, root & tuber crops, mushrooms, nuts & seeds, fruits, fish & shellfish, sea algae, and meat. Given that the food intake rate is not limited to the air fryer cooking method, this study may have been overestimated. Also, many factors, such as country and diet, can affect dietary exposure. Nonetheless, this result indicates that when consuming food cooked with an air fryer, people in this age group need to consider dietary exposure to AA.

In summary, this study investigated the presence of toxic contaminants (PAHs, AA, and HAAs) in food cooked using an air fryer. The concentrations of PAH4 and HAA8 were observed to be lower than those associated with traditional cooking methods. However, specific food groups exhibited higher concentrations of AA. A significantly negative relationship was identified between PAH4 and HAA8 production. Air-fried foods were classified as having ‘low concern’ for PAH4 and PhIP, but the average dietary exposure to AA in cereals, vegetables, and tuber crops might be a concern (below 10,000 MOE values). Most food categories in P95 dietary exposure to AA were flagged as ‘may be a concern’ underscoring the importance of minimizing AA in air-fried foods, particularly those known to contain significant AA quantities.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Conflict of interest

All authors declare that there are no conflicts of interest.